Raw data

29 out of 51 studies reported an outcome of Fever. All grades combined.

Treatment groupings for analysis

Consider four alternative groupings from finer to coarser:

  1. Bisphosphonate drug, dose and delivery method. IV placebo, oral placebo and observation only distinguished.

  2. Bisphosphonate drug and delivery method.

  3. Bisphosphonate drug. IV and oral placebo grouped, but distinguished from observation only.

  4. Drug class, considering all nitrogenous bisphosphonates together. IV and oral placebo grouped, but distinguished from observation only.

Notes

  • Only study including Clodronate was NSABP B-34, whose control arm used oral placebo. 1 and 0 people with fever were reported in the active / control arms respectively, out of 1623, 1612. No other study used oral placebos. Exclude for now, but may be included later if we decide to combine all placebos or all controls.

  • Two studies, ABCSG12 and HOBOE, included separate arms to compare a different kind of hormone therapy for the same bisphosphonate (“main” treatment). To start with, define these as different treatments.

Networks of comparisons

Number of studies for each comparison indicated on the connecting lines.

Network meta analysis models

Model assumptions

  • Relative effects for each comparison are “random effects”, normally distributed between studies with an unknown mean and variance

  • Baseline rate of adverse events (that is, for “observation only” controls) is also a random effect. Log odds is normally distributed between studies with an unknown mean and variance.

  • gemtc R package used, with code modified to deal with random baselines. Package default weakly informative priors.

The only relative effects which can’t be estimated from the data are those relating to Ibandronate. There is only one study for this drug (Body 2007) which reported data for fever. 26 out of 137 in the zoledronic acid arm reported fever, compared to 0 / 137 in the ibandronate arm. Unsure how to deal with this! Is it clinically plausible? We might be able to moderate this outlying result using some model which borrows information from other drugs which are similar to ibandronate? Or just merge all nitrogenous bisphosphonates?

## Loading gemtc

Relative treatment effects

Estimates from each of the four network meta-analysis models, compared with direct data, for the comparisons where there is direct data.

The four alternative ways of grouping effects of similar treatments appear to give a similar fit to the data, according to standard methods of statistical model comparison (DIC)

Absolute rates of fever

To produce the absolute event rate under each treatment:

TODO did I forget to sample the random treatment effects. does this even matter? haven’t illustrated heterogeneity in treatment effects

Posterior mean and 95% credible intervals

Influence of additional treatment

The figure below illustrates the event rate under “observation only” for each study, with studies grouped in the plot according to the additional treatment(s) given. The network meta-analysis model was used, so that estimates could be obtained for studies which did not include an “observation only” control group.

So we could calculate predicted baseline event rates either by

Between-study heterogeneity in relative effects

Limited amount of information to investigate between-study heterogeneity in the relative treatment effects. There are 6 studies each which compared Zol 1 IV and observation, and Zol 2 IV and observation. Estimated odds ratios from these are plotted below.

Heterogeneity here seems driven by small counts, unlikely to be enough information to determine any predictors of heterogeneity.

Each other direct comparison has at most 2 studies (see network plots above).

TODO